Analysis of questionnaires from the perspective of hospitalization risk prediction. Systematic review

Abstract

One of the key tasks facing the medical community and health professionals is the identification of patients at high risk of hospitalization, including emergency one. Currently, various approaches are used to obtain information about certain predictors in a patient that are as­sociated with an increased hospitalization risk, but there is no consensus on which of them are most effective.Aim. To summarize data from studies that has analyzed the prognostic value of various predictors used in questionnaires or patient surveys associated with an increased hospitalization risk.Material and methods. We search for full-text articles in 8 Russian and foreign electronic databases (Scopus, PubMed, BMC Health Services Research, BMJ Research, Google Scholar, e-library, Oxford Academic, Wiley Online Library), published in the period from January 1, 1993 to January 1, 2023 and containing information about questionnaires on hospitalization risk prediction. The study was carried out in 4 following stages: 1 — search for articles using keywords in electronic databases; 2 — duplicate elimination; 3 — search for full-text versions of articles and their assessment for compliance with inclusion/ exclusion criteria; 4 — data analysis, systematization of hospitalization predictors used in questionnaires.Results. In total, out of 28133 initial publications, 13 full-text papers describing studies that fully met the inclusion/exclusion criteria were admitted to stage 4 and were included in the final analysis. These studies were devoted to the description of 11 questionnaires, their modification or testing. The majority of the articles found (53,8%) were conducted in the United States. When analyzing questionnaires to assess the hospitalization risk, among 61 identified predictors, the following were most often used In questionnaires: "age ≥75 years"; "≥1 hospitalization within one to two years preceding filling out the questionnaire"; "taking ≥5 prescription medications"; "male sex"; "history of diabetes", and "satisfactory/unsatisfactory self-perceived health status". A more detailed analysis of questionnaires to assess the risk of one-year hospitalization among people ≥65 years old found that the following predictors were most often used: "age ≥75 years", "≥1 hospitalization in the period from one to two years before filling out the questionnaire", and "taking ≥3 prescription medications".Conclusion. Evidence is summarized on predictors used in questionnaires on hospitalization risk assessment. Additional research is needed to validate established predictors in the Russian population, their cause-and-effect relationship, and to search for novel predictors influencing the hospitalization risk among the working age and post-working age population.